The article is devoted to the development of various approaches to web application development based on machine learning models. The goal is to create intelligent web applications. Each page that does not comply with the most important rules of web application design will contain annoying, intrusive ads (especially if they are irrelevant to the user), which will slow down the workflow. With such a large choice in the IT market, users face great difficulties in choosing between them. The key strategy for intelligent web applications should be the immediate integration of machine learning model rules into the application design. Special approaches must be developed to create such rules. The article provides information on supervised learning models and their capabilities for creating intelligent web applications. The types of intelligent web applications and methods for developing their schemes, as well as the results of their testing, are analyzed. Given this, the main problem outlined in the article is the selection of machine learning models and the development of various approaches to them to create convenient and intelligent web applications for network users. Therefore, for nonlinear supervised learning problems, machine learning models such as K-Nearest Neighbors, K-Means, Support Vector Machines, Decision Trees, Gradient Boosted Machines, Random Forests are analyzed. Of these, K-Means, Decision Trees, and Random Forests models have been selected and different approaches have been developed for them.